Accelerator-based Validation of Shielding Codes C. Zeitlin1, L. Heilbronn1, J. Miller1, J. W. Wilson2 1. Lawrence Berkeley National Laboratory 2. NASA Langley Research Center Abstract The space radiation environment poses risks to astronaut health from a diverse set of sources, ranging from low-energy protons and electrons to highly-charged, high-energy atomic nuclei and their associated fragmentation products, including neutrons. The low-energy protons and electrons are the source of most of the radiation dose to Shuttle and ISS crews, while the more energetic particles that comprise the Galactic Cosmic Radiation (protons, He, and heavier nuclei up to Fe) will be the dominant source for crews on long-duration missions outside the earth’s magnetic field. Because of this diversity of sources, a broad ground-based experimental effort is required to validate the transport and shielding calculations used to predict doses and dose-equivalents under various mission scenarios. The experimental program of the LBNL group, described here, focuses principally on measurements of charged particle and neutron production in high-energy heavy-ion fragmentation. Other aspects of the program include measurements of the shielding provided by candidate spacesuit materials against low-energy protons (particularly relevant to extra-vehicular activities in low-earth orbit), and the depth-dose relations in tissue for higher-energy protons. The heavy-ion experiments are performed at the Brookhaven National Laboratory’s Alternating Gradient Synchrotron and the Heavy-Ion Medical Accelerator in Chiba in Japan. Proton experiments are performed at the Lawrence Berkeley National Laboratory’s 88” Cyclotron with a 55 MeV beam, and at the Loma Linda University Proton Facility with 100 to 250 MeV beam energies. The experimental results are an important component of the overall shielding program, as they allow for simple, well-controlled tests of the models developed to handle the more complex radiation environment in space. Introduction Three broad categories of space radiation are simulated in experiments performed by our group at ground-based particle accelerators: High-energy protons and ions (HZE) in the Galactic Cosmic Radiation (GCR); medium-energy protons from the GCR and from Solar Particle Events (SPE); and low-energy trapped protons. In terms of absorbed dose, these are the most important particle types encountered by humans in space. GCR-like particle beams can be obtained at the Alternating Gradient Synchrotron (AGS) at the Brookhaven National Laboratory and at the Heavy Ion Medical Accelerator in Chiba (HIMAC) at the National Institute of Radiological Sciences, Japan; medium-energy protons are available at the Loma Linda University Proton Treatment Center; and lowenergy protons at energies approximating trapped protons are produced at the Lawrence Berkeley National Laboratory’s 88" Cyclotron. In all of these categories, experimental data obtained by our group provide checks on model calculations, particularly those of the NASA-LaRC group, whose codes are used in spacecraft and spacesuit shielding models. The experiments and calculations described here fall under two headings: fundamental physics relevant to the development of transport codes, and shielding materials studies. In the former category, we include measurements of cross sections for charged-particle production in HZE reactions [1, 2, 3] and cross sections for neutron production in HZE reactions [4], using elemental targets in both cases. (The experiments for measuring the two types of cross sections are quite distinct from one another, as described in the references. Here, we will focus on the charged-particle results.) The data obtained in these experiments are used for comparison to transport models, for example, the NASA-LaRC

code NUCFRG2 [5], QMSFRG [6], and others. The ultimate aim of the measurements is to fill in many of the holes in the existing database of cross sections for nucleus-nucleus interactions, which should lead to improvements in the semi-empirical portions of the transport codes. In the second category of experiments, the shielding materials studies, we include several experiments: iron-beam fragmentation measurements using composite targets, performed primarily at the AGS [7]; transport of energetic protons (150 to 250 MeV) at LLU using a human phantom placed inside the EMU spacesuit used by U.S. astronauts during extra-vehicular activities [8]; and the simulation of low-energy trapped protons traversing thin layers of spacesuit fabric candidate materials, performed at the LBNL 88” Cyclotron [9]. Here, owing to space limitations, we concentrate on the studies using GCR-like beams: the charged-particle cross section measurements and the composite target data from the AGS. Charged-particle Cross Sections in HZE Reactions In these experiments, HZE ions at GCR-like energies are produced at the AGS and HIMAC, and elemental targets are placed in the path of the beam. The targets used are C, Al, Cu, Sn, Pb, and, in some recent experiments, a Li target has been used. Polyethylene (CH2) is also; when cross sections obtained with the C target are subtracted from those with polyethylene, one obtains the cross sections on H, which are of particular interest for two reasons. First, and most relevant here, calculations [10] predict that hydrogenous materials provide superior shielding. Second, pertinent to astrophysics, the observed GCR flux is the result of transport from the sources through the interstellar medium, which is primarily hydrogen. Thus, accurate measurements of H-target cross sections are needed if the GCR composition at the source is to be understood. The experiments described here use silicon detectors placed along beam axis (0°). Detectors are placed so as to subtend varying acceptance angles, as illustrated schematically in Fig. 1. Many data sets have been obtained; Table 1 summarizes the data sets and the status of the analysis. In addition to the 0º data, a few data sets have been obtained with detectors placed off the beam axis, in order to make direct measurements of fragment angular distributions. The existing data on angular distributions are quite sparse. target Scint 1 d3mm3/4

d3mm1/2

Scint 2 TOF path 2-4 m

d3mm5/6 NaI

Fig. 1. Schematic layout of charged-particle experiments at HIMAC. At the AGS, the TOF (time-of-flight) is not used due to the limited space available along the beamline. Spectra obtained at 0° are used to extract the charge-changing and fragment production cross sections (σcc and σZ), which depend on the species of the beam and target nuclei, and on the beam energy. Over the range of energies pertinent to the GCR, energy

# events per 0.1 charge unit

dependence of the cross sections is observed to be mild. As shown in Figure 2, significant differences are seen in spectra obtained with detectors placed at large acceptance (i.e., near the target exit, as in Fig. 2a) compared to spectra obtained with detectors further downstream (Fig. 2b). These differences give indirect information about fragment angular distributions. We note that virtually all previously published 0° measurements used only large acceptance detectors, and did not report cross sections for light fragments.

(a) Large acceptance

(b) Small acceptance

Fig. 2. Spectra obtained with the 600 MeV/nucleon 28 Si beam at HIMAC incident on a 1.31 g cm-2 carbon target.

The spectrum in Fig. 2a illustrates the reason for the lack of light-fragment cross sections from earlier experiments: at large acceptance, there are simply no well-resolved fragment peaks below approximately half the beam charge, making particle identification at the low end impossible. This is a consequence of the relatively high multiplicity of fragments produced when the incident beam ion undergoes a central collision and is broken up into many lighter nuclei. Because light fragments have relatively broad angular distributions, many of them are outside the acceptance of detectors placed far downstream of the target (i.e., at small acceptance). The reduction in observed multiplicity produces very clean spectra like the one in Fig. 2b. Note that although the fragment multiplicity is reduced at small acceptance angles, there are nonetheless several peaks in Fig. 2b that can only be explained by the presence of at least two fragments detected in coincidence, e.g., the peak near Z = 3.5, which is likely due to events in which three He fragments are detected. Other multiple-fragment peaks appear as shoulders just above the singlefragment peaks at 4, 5, 6, and 7. Another notable feature of Fig. 2b is the peak of approximately 3000 events (out of about 300000 total) at Z = 0; these are events in which all the charged fragments were outside the acceptance of this detector. Once the Z spectra are produced, a simple count of the numbers of events detected as a function of species is the main input to the calculation of the charge-changing and fragment production cross sections. Fig. 2 shows that even-charged fragments are preferentially produced, resulting in enhanced fragment production cross sections for those species, as illustrated in Fig. 3. The production of F (Z = 9) is particularly suppressed. Details such as these provide sensitive tests of models of nuclear interactions.

Si + C, 600 MeV/nucleon

Fig. 3. Fragment production cross sections for 28Si on carbon at 600 MeV/nucleon, showing the “odd-even” effect (enhanced production of evenZ fragments).

σ (mb)

160 120 80 40 0 5

7

9

11

13

Fragment Z

Charged-particle Fluences Behind Thick Targets In a series of accelerator experiments similar to the cross-section measurements, we have placed thick, usually composite targets in the path of heavy ions beams. Most of our data have been acquired with the 1087 MeV/nucleon 56Fe beam at the AGS. Fluence spectra have been obtained for many targets. Because the beam energy is relatively high, effects of fragmentation predominate over ionization energy-loss effects. Thus, in this beam, for any target material, the dose per particle exiting the target will be decreased relative to the dose per particle of the incident beam ions, provided the areal density is less than about 20 g cm-2. (Especially at lower energies, there is a tradeoff between the two: Energy loss tends to increase LET and hence dose per particle, for those particles with sufficient energy to traverse the full depth of the target; fragmentation reactions, on the other hand, reduce the nuclear charge, reduces the dose per particle.) Despite this general trend, variations are still seen when we quantify the reduction in dose per particle reduction with different shielding materials. The analysis of these data is simple. Starting with the LET of the 1087 MeV/nucleon AGS 56Fe beam (148 keV/µm), using target-out data we determine a scale factor to convert the measured quantity (deposited energy, ∆E, in silicon) to LET. This factor is then applied to convert ∆E distributions acquired with targets in the beam to LET spectra. In a manner similar to that employed in the cross-section data analysis, we obtain counts of fragments and surviving primaries by species and apply (small) corrections for losses in the detector. The corrected counts are then used to obtain a corrected average LET, , which with this beam is invariably lower than the beam LET, Lbeam. Because the targets are of modest depth and the beam energy is high, there are essentially no particles stopped in the target, and therefore (in the simplest approximation), the fluence exiting the target is equal to the incident fluence. Therefore is directly proportional to the dose of the mixed field seen behind the target. For each target, we compute the fractional dose reduction per unit mass, given by the simple relation f = (1 - / Lbeam)/(ρ*x) where ρ*x is the areal density of the target. Results are shown in Fig. 4.

Ion Species 4 He 12 C 12 C 14 N, 16O 20 Ne 28 Si 40 Ar 56 Fe 56 Fe 56 Fe 56 Fe

Beam Energy (MeV/nucleon) 230 290 400 400 600 400, 600, 1200 400 400 500 600 1087

Accelerator HIMAC HIMAC HIMAC HIMAC HIMAC HIMAC, AGS HIMAC HIMAC HIMAC AGS AGS

Status 1 3 3 2 4 3 1 2 1 2 4

Table 1 Heavy-ion data sets obtained through August 2002. Status key (as of 8/02): 1 – not yet analyzed; 2 – analysis in progress; 3 – analysis completed, manuscript in preparation; 4 – analysis completed, published. 1 .0 6 G e V /a m u 5 6 F e P R E L IM IN A R Y -2

frac tion d os e re du ctio n p e r g

Cu

4 .0 E -0 2

Al P E T I-5

3 .5 E -0 2

Al-C H2-7 C H2/L i6

3 .0 E -0 2

Al-C H2-1 0 LiF

2 .5 E -0 2

C H2/L iC C H2*

2 .0 E -0 2 1 .5 E -0 2 1 .0 E -0 2 5 .0 E -0 3 0 . 0 E +0 0

C H2+ B C H2 G r- E pox10 P ur e C G r- E pox5 Mars Reg olith B -E po xy E poxy

Fig. 4. Dose reductions for various materials placed in the 1087 MeV/nucleon 56Fe beam at the AGS. The list of target names in the legend, starting at the top, corresponds to the entries in the chart from left to right (i.e., the left-most chart entry is for the Cu target, followed by Al, PETI-5, etc.). In [10], Wilson et al. predict that hydrogenous materials will, in the space radiation environment, give the best performance per unit mass in terms of both dose equivalent and cell transformation probability. The results in Fig. 4, though not directly comparable, are qualitatively consistent with this prediction: the greatest dose reductions are seen with the

Epoxy and Boron-Epoxy targets, which are hydrogen-rich. The Martian regolith simulant target also does well; it too has considerable hydrogen content (it consists of polyethylene blended with a mixture that approximates Martian soil). The smallest dose reductions are seen for elemental targets, Cu and Al (Pb data are also available but have not yet been analyzed for this purpose). Many more data sets have been obtained that can be applied to this analysis. Varying the beam ion and energy will provide a broader test of the relative performance of the different materials. Summary Extensive charged-particle measurements with GCR-like ions have been made for purposes of improving the nuclear cross section database and for direct measurements of shielding properties of various materials. Though considerable progress has been made in recent years, a large archive of data still awaits analysis and publication. References 1. C. Zeitlin, L. Heilbronn, J. Miller, S. E. Rademacher, T. Borak, T. R. Carter, K. A. Frankel, W. Schimmerling, and C. E. Stronach, Phys. Rev. C 56, 388 (1997). 2. C. Zeitlin, A. Fukumura, L. Heilbronn, Y. Iwata, J. Miller, T. Murakami, Phys. Rev. C 64:024902 (2001). 3. C. Zeitlin, A. Fukumura, L. Heilbronn, Y. Iwata, J. Miller, T. Murakami, LBNL Report Number 47655 (2002). 4. Y. Iwata, T. Murakami, H. Sato, H. Iwase, T. Nakamura, T. Kurosawa, L. Heilbronn, R. M. Ronningen, K. Ieki, Y. Tozawa, K. Niita, Phys.Rev. C 64:054609 (2001). 5. J. W. Wilson, J. L. Shinn, L. W. Townsend, R. K. Tripathi, F. F. Badavi, S. Y. Chun, Nucl. Instr. Meth. B 94, 95 (1994). 6. F. A. Cucinotta, J. W. Wilson, R. K. Tripathi, L. W. Townsend, Adv. Space Res., 22, 533 (1998). 7. J. Miller, C. Zeitlin, L. Heilbronn, F. A. Cucinotta, LBNL Report Number 49423 (2002). Accepted for publication in Radiation Research. 8. C. Zeitlin, L. Heilbronn, J. Miller, M. Shavers, LBNL Report Number 47656 (2001). This report is available on the World Wide Web at: http://www-library.lbl.gov/docs/LBNL/476/56/PDF/LBNL-47656.pdf 9. J. W. Wilson, J. Tweed, C. Zeitlin, M-H. Y. Kim, B. M. Anderson, F. A. Cucinotta, J. Ware, A. E. Persans, 31st International Conference on Environmental Systems, Orlando, Florida, July 9-12, 2001. 10. J. W. Wilson, M. Kim, W, Schimmerling, F. F. Badavi, S. A. Thibeault, F. A. Cucinotta, J. L. Shinn, and R. Kiefer, Health Phys. 68, 50 (1995).